Coincidentally, About 5,175 Words

L. Timmel Duchamp recently linked to and discussed a post by Jonathan McCalmont entitled "Annoyed With The History Of Science Fiction." It is (as Duchamp says) a very useful rant inspired by yet another post, this one a naïve and frankly unnecessary reappraisal of Robert A. Heinlein by Gary Westfahl. McCalmont's subject is not so much Westfahl's arguments regarding Heinlein as the broader critical naïveté in the sf field which Westfahl's essay, to McCalmont, represents, and a call for a more sophisticated, "technical" sf criticism. The problem, he says, is that too much sf criticism relies on simplistic historical approaches and catalogs of plot synopses, resulting in a so-called criticism that is more an accumulation of half-baked, unsupported assertions and superficial natterings about personality and "influence" than any kind of attempt to come to terms with the texts at hand--and their relationships to the world. A better approach is emblematized for McCalmont--and me, and, it would appear, Duchamp--by Samuel R. Delany's famous essay "About 5,175 Words," which, McCalmont argues, issued a vital call for more that was, for all the attention the essay received and continues to receive, never particularly answered.

McCalmont's essay and Duchamp's brief but provocative response (in which she issues her own call for more analysis of sf's "special, particular aesthetics and sensibility," particularly as it is currently changing, and as relates to feminist sf) both contain a wealth of important material. I was immensely excited to see them, as much in them seemed to cut right to the heart of what I've been feeling about sf, but have only rarely seen discussed. Most exciting of all, for me, is that they have given me a point of entry to get into a discussion of all these issues, which until now I have been struggling to find. There is more going on here than I will be able to cover in a single post, no matter how long (and I do tend to go on), and so I hope to begin here a series of essays prompted by these posts (and perhaps, to a lesser extent, the original Westfahl essay that started the whole thing, as it in its own way brings up some very important points). Topics for future posts will hopefully include (probably some but not all of) the following: a discussion of "sense of wonder," which Duchamp understandably but I think wrongly dismisses in her post; an exploration of some aesthetic issues particular to feminist sf (a necessarily rudimentary exploration, given my apparently incurable dudeness and the still woeful state of my reading in this area); the continued, systematic marginalization of the voices of women and other marginalized people within the sf field; the stylistic and aesthetic features of sf by virtue of which it nevertheless possesses considerable value for these marginalized populations; my difficulties with contemporary sf; and a number of other aesthetic and social issues raised by these extraordinary posts; along with, I hope, many more such issues not raised in them but which I have been considering for some time now. For now, though, I would like to focus on a section of McCalmont's post which Duchamp also highlights, in order to enter into a discussion of the implications of methods of exposition in sf.

McCalmont objects to Westfahl's assertion of Heinlein's influence on the sf field for many reasons which need not concern us at the moment; but what he feels is particularly "frustrating" is that this influence is

not only taken for granted but assumed to be positive. For example, the received opinion is that Robert Heinlein pioneered a number of techniques that are now used widely within the field but when you attempt to ascertain what these techniques might have been you will struggle to find anything more involved than an airy assertion that Heinlein's fondness for sentences such as "the door irised open" marked a radical improvement over the field's historical reliance upon a form of lead-footed exposition now dismissively referred to as 'info-dumping'.

Terms like 'info-dumping' are the science fiction equivalent of the film critic's 'deep focus', 'long take' and 'dynamic editing'. However, while film critics are able to draw upon a rich technical lexicon, the few technical terms used by SF critics generally come bundled up with their own unexamined assumptions about how best to write science fiction. For example, the lionisation of show-don't-tell at the expense of the info-dump assumes that the aim of science fiction is to tell a story that is immersive in that it never causes the reader to break from the story and think about what they have just read.

My own critical interest in the style of sf arose originally from my attempts to understand, both as a reader and as a prospective writer, why it was that writing typically described, even by many of its devotees,* as "clunky" or even just "bad" could have such a profoundly moving impact on me; and McCalmont's comments regarding the infodump,** which to a certain extent mirror many of my own thoughts, are crucial to this enterprise.

*Among whom a common, and to me infuriating and nonsensical, formulation is "They're a terrible writer, but their ideas are great."
**I prefer to write it without the hyphen, which is probably an American thing; your mileage, as they say on the internet, may vary.

Before I go further, a word about terms. To begin with, McCalmont is entirely correct when he says that the term "infodump" comes laden with a lot of negative baggage; however, I see no reason why a hopefully more sophisticated critical language cannot rehabilitate it. It has the advantage of being nearly universally understood among sf readers (i.e., though most readers might assume reflexively that it is a "bad thing," they will all at least know what you mean when you say it), beyond which, frankly, I like it as a word, not least because it really does convey an essential aspect of the feeling one gets on actually encountering an infodump in a work of fiction. So I will use it throughout, in the understanding that it is to be taken as a value-neutral term: indicating only the technique of placing passages of straight exposition into a narrative, implying no assessment of the validity of this technique--beyond, of course, what I state explicitly. Next, though McCalmont uses the familiar phrase "show-don't-tell" (and I shudder, as I imagine he may have, merely typing it) to describe what those who oppose the infodump suggest should be used in its place, to me this is inadequate. Mostly this is because show-don't-tell is a technique typically advised not only for exposition but for "action" as well.* As we are speaking specifically of expositional techniques--and of expositional techniques unique to sf--a more particular term seems desirable to me, and an appealing one I've picked up from somewhere (I have a vague impression I might have first picked it up from Jo Walton) is "incluing": that is to say, the process of gradually "cluing the reader in," indeed by showing-not-telling, revealing information through action which the reader then pieces together as they read to form an evolving picture of the world of the story.** Thus, Heinlein's famous "The door dilated" is incluing, where if he had written something along the lines of "The door was constructed out of a number of separate panels, which by a mechanism were made to etc.," it would have been infodumping.

*Here I find I am running into a problem of terminology myself. To a large degree, every word in a piece of fiction is "exposition," in that it is these words that give us the information out of which we construct the story--and in discussion of non-sf fiction this sense of the word is often the most fruitful. But in this discussion, and in general all discussions of sf, with its highly particular expositional problems, it is useful to distinguish between narrative action--the "what is happening" of the story--and what I will here refer to as exposition, by which I mean the relation of information which we are meant to take as existing in the imaginary space outside of the story, but which is being brought into the story for the purpose (usually) of increasing our understanding of the events of the story. As I am about to discuss, the two goals can be achieved simultaneously, but even in these cases they are worth distinguishing as, to some extent, different goals. There are, I am aware, many problems and limitations in this model, but I think it will serve well enough as a starting point.
**Another advantage of the term "incluing" is that it avoids the tendentious and frankly nonsensical implication (which I am not, to be clear, imputing to McCalmont) that it is ever possible, in the medium of the pure written word, to "show" the reader anything rather than to tell them, and thus in itself contains, I hope, none of the naïve critical assumptions I am trying to avoid. Additionally, the word's formation is roughly parallel to that of infodump, which is nice.

There is, I think, to some extent a general understanding among readers, writers, and critics of sf that these are the two primary methods of exposition, and that the central importance of exposition in sf means that these techniques are themselves important. However, understanding and discussion of what the techniques actually are, and more importantly what they actually do, seems generally limited to matters of preference: incluing as "good writing," infodumping as "bad writing," for example, with little discussion of why this might be other than appeals to the supposed virtue, as McCalmont mentions, of a smoothly flowing narrative in which readers can lose themselves.

McCalmont clearly thinks that this is not in itself a virtue, and though all my sympathies are with him in this matter, I find that I cannot agree with his implication (whether it is what he intends or not) that the preference for incluing over infodumping merely needs to be reversed. I think that what is first needed is an examination of what these techniques do, what effect they have upon the experience of reading--and it is my belief that such an examination will reveal that, while the two techniques are indeed vastly different, the opposition between the two is illusory and, indeed, untenable; and that what is needed is not an argument over which is the more valid or useful or beautiful or "realistic" or whatever value-oriented adjective one wishes to apply, but rather an appreciation of the formal/structural issues involved in their deployment: that is, the reasons why a writer may choose here to inclue, here to infodump, and what impact these choices have on the reader.

In Gabriel Josipovici's What Ever Happened to Modernism?, a work surprisingly applicable to the study of sf (and every bit as central to my own current understanding of it as the critical works of Delany and Russ), he says of Robert Pinget's Passacaille:

The narrative is both much slower and much faster than in a traditional novel. Like a piece of music by Birtwistle it spirals forward via repetitions which are never quite repetitions, until we find ourselves in possession of far more information than would have been the case in a conventional narrative or symphony... There is this insistent counterpoint to the detail... The narrative goes calmly on its odd way, as more and more elements are dropped in... [T]he book leaves one with the sense of having participated in the birth of narrative itself...

Our response to it...is likely to be precisely that which Malcolm Bowie posited [in Mallarmé and the Art of Being Difficult] of the reader of a Mallarmé poem: either to try to get 'with panic stricken rapidity' at 'what it means', or to abandon it for ever. 'The double effort required to allow Mallarmé's gaps their full disjunctive power', we recall Bowie saying, 'yet at the same time remain attentive to the multitude of invisible currents which pass back and forth between the separated segments, will strike many readers as inexcusably arduous and unrewarding.' Yet, he concludes, 'the view I shall propose is that time spent learning to read Mallarmé is amply repaid.' I would only add that Bowie is perhaps a little too defensive, or at least that reading Pinget, Simon or Robbe-Grillet is infinitely easier than reading Mallarmé, and that it is exhilarating rather than arduous. But then I imagine Bowie really believes this holds true for Mallarmé as well.

Now, the techniques of Pinget and of Mallarmé are, clearly, enormously different from those of most sf writers (and from one another), and where with the modernists Bowie is correct to speak of the perceived "arduous and unrewarding" nature of the works, and of the value of "time spent learning to read" them, in sf we might replace these phrases with, respectively, "clunky and simplistic" and "giving the writer the benefit of the doubt" (the latter of which in many ways amounts to the same thing). But for me, despite these immense differences in surface-level style and affect, there is a striking resonance between Josipovici's and Bowie's analysis here and the feeling I get from sf novels.

As so often happens, I find that I am having difficulty going on, because my thoughts on these matters shoot out in so many directions simultaneously; all I can do is ask that you try to bear with me as I attempt to put them into some kind of order.

For me, one of the things that fundamentally makes sf, sf, is that it goes out of its way to require more exposition than other literary modes. One cannot just "tell a story" in sf; one must also carefully and complicatedly establish the world in which that story "takes place" in order for the story to be in any way understandable. When you think about it, this is actually quite astonishing, especially considering that it is commonly agreed, or at least commonly asserted, that sf grew out of and to a certain extent remains "popular adventure fiction," in which the straightforward relation of incident, one would presume, should be paramount. But instead, sf by its very nature frustrates the reader's continual desire to "find out what happens next." Is it not truly remarkable that, from the very beginning (wherever one places this: with Gernsback and Amazing, with H.G. Wells or Jules Verne, with Mary Shelley, what have you), we find sf stories repeatedly stopping dead in their tracks while the narration lectures the reader on various matters of fact, real or imagined?

Looked at this way, we could perhaps say that a strength of the infodump is that it delays narrative. It encourages us to stop allowing ourselves to be swept away in fictional incident, to look up from the book and think about what we are doing by reading, what is being done to us, and to examine the difference between the fictional world we are building up in our minds and the world in which we live--in a way that a typical realistic novel never does (indeed, cannot afford to), but very similar to the way that modernist novels do. More, infodumps sometimes allow the stories that contain them to become boring for a time, and here it is perhaps useful to recall Susan Sontag's comment* that boredom can be a valid literary technique, analogous in some ways to the use of silence in music--which use can range from simple, but powerful, punctuation, as in "The Little Girl I Once Knew" by The Beach Boys, to a radical effort to recontextualize the everyday world outside of the artwork by allowing consideration of it to become part of the work, as in 4'33" and other compositions of John Cage.

*Which I unfortunately do not have to hand, but I believe that it is in the essay "On Style," collected in Against Interpretation.

In a way, the notion that the infodump is somehow a "primitive" technique which has been superseded by the more "sophisticated" inclue is reminiscent of the commonly held notion that the chorus of ancient Greek tragedy was a primitive device, and that the reduction of its importance in Euripides and its elimination entire by later playwrights was a progressive innovation to be admired as such, no more, no less. Both narratives rely on an assumption of teleological progress which treats the past as no more than some sort of a trial run for the present, which is parochially assumed to be both universally "better" and the only way that things could have "ended up." In the process, not only the techniques seen as belonging to the past but also those of the present are underestimated, treated not as the responses of artists to the pressures and needs of their worlds but rather as steps in some process of objective "improvement," as though, say, the chorus-free drama was in beta when Euripides wrote (or the infodump-free sf, perhaps under John W. Campbell?), and after the bugs were worked out it was ready for full release later on. In both cases, the actual experience of the audience of the time--past or present--not to mention that of the individual artist, is lost.

What, then, of the inclue? Well, just because I think the "evolution" away from the infodump is not something with which we should be pleased (and that its continued gleeful use by writers who ignore misguided advice should not be denigrated), it does not mean that the inclue is by some see-saw motion necessarily a "bad thing." I doubt that this is what McCalmont thinks, either, or at least not in any strong sense; however, there is a strong implication in his essay that the primary use of incluing is to create the kind of "immersive" narratives that he suggests should not be the goal of sf--as when, after mentioning Neal Stephenson, Kim Stanley Robinson, and Stanislaw Lem as examples of writers who "make frequent use of info-dumps as they believe that wading through densely written expositional text is an integral part of the science fiction experience," he goes on to suggest "that Lem's approach to info-dumping is so effective and idiosyncratic that it not only forms an integral part of his novels' literary affect, it also makes his work substantially more complex and interesting than anything written under the purview of show-don't-tell," or what I am referring to as incluing. His comments on infodumping and these author's uses of it are, as I have said, very important, and I agree wholeheartedly with them; where I differ is in the opposition McCalmont sets up between infodumping and incluing, and the notion that incluing is, or can only be, contributory to simplistic "immersion."

As McCalmont makes reference to both Heinlein's sentence "The door dilated" and the critical works of Delany, it will not be too much of a leap to go to Delany's remarks on this very sentence. In that same essay which McCalmont praises, "About 5,175 Words" (or, in the revised version that appears in the latest edition of The Jewel-Hinged Jaw, from which I quote it, "About 5,750 Words"), Delany quotes Harlan Ellison on Heinlein's sentence--and for once, Ellison has something valuable to say:

...Heinlein has always managed to indicate the greater strangeness of a culture with the most casually dropped-in reference: the first time in a novel, I believe it was in Beyond This Horizon, that a character came through a door that...dilated. And no discussion. Just: "The door dilated." I read across it, and was two lines down before I realized what the image had been, what the words had called forth. A dilating door. It didn't open, it irised! Dear God, now I knew I was in a futuristic world...

(All italics and ellipses are present in Delany.)

Delany, bringing to bear his concept of sf's subjunctivity, comments that the sentence "is meaningless as naturalistic fiction...As SF--as an event that hasn't happened, yet still must be interpreted in terms of the physically explainable--it is quite as wondrous as Ellison feels it." Elsewhere, at "Shadows" 38, Delany has more to say about this sentence:

Science fiction is science fiction because various bits of technological discourse (real, speculative, or pseudo)--that is to say the "science"--are used to redeem various other sentences from the merely metaphorical, or even the meaningless, for denotative description/presentation of incident. Sometimes, as with the sentence "The door dilated," from Heinlein's Beyond This Horizon, the technological discourse that redeems it--in this case, discourse on the engineering of large-size iris apertures, and the sociological discourse on what such a technology would suggest about the entire culture--is not explicit in the text. Is it, then, implicit in the textus? All we can say for certain is that, embedded in the textus of anyone who can read the sentence properly, are those emblems by which they could recognize such discourse were it manifested to them in some explicit text.

There is much to explore here. A good starting point might be to examine the experience Ellison reports with Heinlein's sentence, which it seems to me safe to assume is fairly representative; I know I, at least, have had very similar experiences with similar sfnal sentences. The key here is that the experience Ellison describes is not one of immersion in the story--rather, it is one of being pulled, abruptly and quite startlingly, out of the story. Even more specifically, it is an experience of having been prepared for immersion, of perhaps expecting or even desiring immersion, to the point where the startling element goes unnoticed...until some process in the mind slower than reading forces one to stop, suddenly, to go back and consider what one has just read--and then to marvel.* At this point, the reader, aware (on whatever level) of the subjunctivity of the text at hand, must construct the various absent discourses to which Delany points in the quote from "Shadows"--otherwise, the sentence must necessarily remain meaningless.

*More on this aspect, I hope, in a future essay on sense of wonder.

A common aspect of naïve discussions of these techniques is the assertion that incluing is "more realistic" than infodumping.* Even putting aside the fact that sf, like modernism, is not a realist literature, and that therefore one would think realism would not even be a criterion for judgment,** this is at best a deeply questionable assertion. It is questionable because, and this tends to get overlooked, what incluing ends up doing is turning the quotidian into a mystery or a surprise. People for whom dilating doors are a part of everyday life would not have anything like Ellison's reaction; they would consider Heinlein's sentence every bit as ordinary as we would consider "The door opened." But for us it is startling; it pulls us for a moment out of the story to consider the differences between the fictional world and our own. More, it forces us to reevaluate every prior appearance of doors in the work, should there be any. At this point, I must admit that I have not read Beyond This Horizon (indeed, I have read very little Heinlein, and have not much cared for what I have read); but let us suppose that the sentence before the one under discussion is something like, "She walked to the door." Leaving aside for the moment any consideration of felicitous writing, I find it remarkable that in a case such as this the bizarre readjustment enforced by the definitively sfnal sentence casts its strangeness back onto the more mundane sentence before it, forcing us to retroactively change our apprehension of that prior sentence.

*This is particularly, but not exclusively, the case when the infodumping takes place in dialogue, specifically in the form often referred to as the "As You Know, Bob" dialogue. This specific form is another topic I hope to discuss at length in the future.
**Admittedly the issue of realism in sf is more complicated than I am allowing here, especially as regards scientific verisimilitude; but that, sigh, is again a topic for some hypothetical future essay.

Another way to put this might be to say that where infodumping, as we have seen, delays narrative (what "happens"), incluing delays knowledge (what "is"). In this connection, we should perhaps keep in mind Josipovici's comments on Pinget. Once again:

The narrative is both much slower and much faster than in a traditional novel... [I]t spirals forward via repetitions which are never quite repetitions, until we find ourselves in possession of far more information than would have been the case in a conventional narrative... The narrative goes calmly on its odd way, as more and more elements are dropped in... [T]he book leaves one with the sense of having participated in the birth of narrative itself...

It would perhaps be taking Josipovici too literally if I were to point in triumph to the word "information" here; and yet I cannot help but feel that, though the methods under examination are radically different, Josipovici's words can apply equally to the best sf novels. The reference to "repetition," too, puts me in mind of Delany's assertion at "Shadows" 37, that "Everything in a science fiction novel should be mentioned at least twice (in at least two different contexts)." Delany leaves this deliberately unexplored, and I will for the moment follow his lead; but I wonder if I am entirely off-base in thinking that this analysis of expositional technique, with the assistance of Josipovici's comments, has taken us quite close to what he means.

Though I do not agree with it, I think I can understand the source of the confusion that leads people to praise incluing as realistic, and others like McCalmont to denigrate it for the same reason. This is because formally, if one leaves aside the transformations wrought upon a text by its being sf, the inclue is very similar to the "telling detail" which is so central to the practice of so-called realist fiction. Josipovici is of much assistance in understanding why readers like McCalmont and myself might be skeptical of the telling detail. Again in What Ever Happened to Modernism?:

The notion that the new reality inhering in novels depends on their attention to detail fails to distinguish between 'reality' and what theoreticians call 'the reality-effect'. In fact [Adam] Thirlwell [whose criticism Josipovici has been discussing as representative of this failure] uses the two terms indiscriminately. But putting a faint scar on a face or alerting us to the fact that the carpet is turned up in the corner, like describing the smell of sweat and semen during the act of sex, no more anchors the novel to 'reality' than writing about stars in the eyes of the beloved. The novel is still made up of words, is still the product of a solitary individual, inventing scars, carpets, smells or stars. Of course we warm to a novelist who surprises us with his attention to detail... Too often though...detail seems to be there as a way of convincing us (and the authors themselves?) that what we are dealing with is the stuff of life.

There is much more in Josipovici's subsequent comments on the telling detail that is extremely relevant to sf, but once more I will have to leave it for another time.* What concerns me now is, as I have said, the superficial similarity between realism's "telling detail" and sf's inclue, which I think has led many commentators astray. But when Heinlein says "The door dilated,"** we are not meant to think that we are faced with reality; rather, we are meant to realize how unlike reality the world of the book is.

*I find myself wishing that everyone had read Josipovici's book, and, more improbable, understood it better than the majority of its professional reviewers, so that I would not feel that I was doing its vitally important arguments a massive disservice by leaving them--for now--largely unmentioned. For the interested who perhaps do not have the time right now to read the whole book (though it isn't all that long!), I recommend Stephen Mitchelmore's review, which admirably summarizes and meditates on the central themes of the book, and blog friend Richard's post reviewing the book and connecting some of the points it raises with more directly political issues gleaned from another essential thinker whose name ends in -ici, the feminist historian and organizer Silvia Federici. Another post of Richard's is relevant in exploring some of the reasons all this "matters."
**And I hope it is clear that I keep returning to this example only out of convenience, for the reason that it is discussed both in McCalmont's essay and in the Delany essays, not because I think it is the last word in masterful incluing. I could just as well be speaking of, say, Delany's casual mentions of the crumpling coffee bulbs in Trouble on Triton, or any number of other examples.

In "Speculations: The Subjunctivity of Science Fiction," which begins as a response to "About 5,175 Words," Joanna Russ speaks characteristically perceptively of the peculiar relationship of reality to sf:

In science fiction the relation between the "secondary universe" of fiction and the actual universe is both implicit and intermittently more or less perceivable. It consists not of what is on the page but in the relation between that and the reader’s knowledge of actuality. It is always shifting.

One does not suspend one’s disbelief in reading science fiction --the suspension of disbelief (complex to begin with, as it is with satire) fluctuates constantly. That is, the relation with actuality--what Delany would call the subjunctivity of the story -–fluctuates constantly.

(I posted a longer excerpt, including this portion, here, and if you have your doubts about her conclusions here, I refer you to the rest.)

The method by which this fluctuation is achieved is precisely the interplay of infodumping and incluing. At times we find that the narrative has been delayed, that the book is reminding us that we are living long before the story it describes takes place (or in a different universe, or what have you), and we perhaps struggle to stay interested, or on the other hand become differently engrossed, in what is suddenly no longer narrative; at other times we find that our knowledge has been delayed, that the book behaves as though we are living in the future (or in the other universe, etc.) with the characters, and we struggle to keep our knowledge at a level which will allow us to function in this unfamiliar world.

All fiction lies. For Josipovici and those of us who take seriously the issues he discusses, fiction must be, and must make the reader, aware of the lie in order to have any chance at approaching anything like truth. To speak overly programmatically and risk eliding the experiences of and differences between individual writers in the interest of brevity (hah!), modernism does this by acknowledging the lie, admitting to it, in any number of ways; sf does it by lying so simultaneously audaciously and rigorously that the lie cannot be ignored, by lying in so complex a way that our awareness of the lie fluctuates, as Russ says, constantly (note too that these methods which I am ascribing separately to modernism and to sf are not mutually exclusive). Whether this works in anything approaching a majority of sf texts is debatable at best, as is the question of whether the average sf author is aware of or cares about any of these issues. But nevertheless it is there--latent, perhaps, but there--in all works of sf; and in the best it is tremendously powerful.